Logo ROOT   6.10/09
Reference Guide
List of all members | Public Member Functions | Private Attributes | Friends | List of all members
TMVA::DNN::Layer Class Reference

Layer defines the layout of a layer.

Layer defines the layout of a specific layer in the DNN Objects of this class don't hold the layer data itself (see class "LayerData")

Definition at line 676 of file NeuralNet.h.

Public Member Functions

 Layer (size_t numNodes, EnumFunction activationFunction, ModeOutputValues eModeOutputValues=ModeOutputValues::DIRECT)
 c'tor for defining a Layer More...
 
std::shared_ptr< std::function< double(double)> > activationFunction () const
 fetch the activation function for this layer More...
 
EnumFunction activationFunctionType () const
 get the activation function type for this layer More...
 
std::shared_ptr< std::function< double(double)> > inverseActivationFunction () const
 fetch the inverse activation function for this layer More...
 
ModeOutputValues modeOutputValues () const
 get the mode-output-value (direct, probabilities) More...
 
void modeOutputValues (ModeOutputValues eModeOutputValues)
 set the mode-output-value More...
 
size_t numNodes () const
 return the number of nodes of this layer More...
 
size_t numWeights (size_t numInputNodes) const
 return the number of weights for this layer (fully connected) More...
 

Private Attributes

std::shared_ptr< std::function< double(double)> > m_activationFunction
 stores the activation function More...
 
EnumFunction m_activationFunctionType
 
ModeOutputValues m_eModeOutputValues
 do the output values of this layer have to be transformed somehow (e.g. to probabilities) or returned as such More...
 
std::shared_ptr< std::function< double(double)> > m_inverseActivationFunction
 stores the inverse activation function More...
 
size_t m_numNodes
 

Friends

class Net
 

#include <TMVA/NeuralNet.h>

Constructor & Destructor Documentation

◆ Layer()

TMVA::DNN::Layer::Layer ( size_t  numNodes,
EnumFunction  activationFunction,
ModeOutputValues  eModeOutputValues = ModeOutputValues::DIRECT 
)

c'tor for defining a Layer

Parameters
itInputBeginindicates the start of the input node vector
itInputEndindicates the end of the input node vector

Definition at line 145 of file NeuralNet.cxx.

Member Function Documentation

◆ activationFunction()

std::shared_ptr<std::function<double(double)> > TMVA::DNN::Layer::activationFunction ( ) const
inline

fetch the activation function for this layer

Definition at line 695 of file NeuralNet.h.

◆ activationFunctionType()

EnumFunction TMVA::DNN::Layer::activationFunctionType ( ) const
inline

get the activation function type for this layer

Definition at line 698 of file NeuralNet.h.

◆ inverseActivationFunction()

std::shared_ptr<std::function<double(double)> > TMVA::DNN::Layer::inverseActivationFunction ( ) const
inline

fetch the inverse activation function for this layer

Definition at line 696 of file NeuralNet.h.

◆ modeOutputValues() [1/2]

ModeOutputValues TMVA::DNN::Layer::modeOutputValues ( ) const
inline

get the mode-output-value (direct, probabilities)

Definition at line 689 of file NeuralNet.h.

◆ modeOutputValues() [2/2]

void TMVA::DNN::Layer::modeOutputValues ( ModeOutputValues  eModeOutputValues)
inline

set the mode-output-value

Definition at line 690 of file NeuralNet.h.

◆ numNodes()

size_t TMVA::DNN::Layer::numNodes ( ) const
inline

return the number of nodes of this layer

Definition at line 692 of file NeuralNet.h.

◆ numWeights()

size_t TMVA::DNN::Layer::numWeights ( size_t  numInputNodes) const
inline

return the number of weights for this layer (fully connected)

Definition at line 693 of file NeuralNet.h.

Friends And Related Function Documentation

◆ Net

friend class Net
friend

Definition at line 712 of file NeuralNet.h.

Member Data Documentation

◆ m_activationFunction

std::shared_ptr<std::function<double(double)> > TMVA::DNN::Layer::m_activationFunction
private

stores the activation function

Definition at line 703 of file NeuralNet.h.

◆ m_activationFunctionType

EnumFunction TMVA::DNN::Layer::m_activationFunctionType
private

Definition at line 710 of file NeuralNet.h.

◆ m_eModeOutputValues

ModeOutputValues TMVA::DNN::Layer::m_eModeOutputValues
private

do the output values of this layer have to be transformed somehow (e.g. to probabilities) or returned as such

Definition at line 709 of file NeuralNet.h.

◆ m_inverseActivationFunction

std::shared_ptr<std::function<double(double)> > TMVA::DNN::Layer::m_inverseActivationFunction
private

stores the inverse activation function

Definition at line 704 of file NeuralNet.h.

◆ m_numNodes

size_t TMVA::DNN::Layer::m_numNodes
private

Definition at line 707 of file NeuralNet.h.


The documentation for this class was generated from the following files: